Training Humans to Teach Robots: Large and Lasting Skill Gains
Yuqing Zhu, Endong Sun, Matthew Howard
AI summary
Problem
Novices struggle to teach robots via learning from demonstration due to misconceptions about robot dynamics, while existing feedback methods lack scalability and generalization.
Approach
The authors developed a scalable training framework that uses machine teaching algorithms to provide real-time, intuitive visual guidance, helping novices internalize control principles that apply to unseen tasks.
Key results
- 75% improvement in novice teaching ability after training
- Teaching gains retained for at least one year post-training
- 71% enhancement in generalizing skills to unseen motor tasks
- Framework validated across simulated and physical robotic platforms
Why it matters
Empowers non-experts to reliably train robots for complex tasks, reducing dependency on expert instructors and accelerating practical human-robot collaboration.
Abstract
Recent evidence has shown that, contrary to expectations, it is difficult for novices to teach robots tasks through learning from demonstration (LfD). Novices often struggle with understanding the relationship between robot states and actions, leading to suboptimal demonstrations. This paper introduces a framework that leverages machine teaching algorithms to train novices in a controlled, ideal environment where optimal control parameters are predefined. The training enables participants to internalise fundamental control princi- ples, preparing them to adapt to new skills that share similar properties. The study evaluates whether such teaching ability is (i) retained beyond the training period (including a long-term follow-up) and (ii) generalised so that novices teach robots more effectively in environments where control parameters are not predefined. It reports a series of between-subjects studies that demonstrate that trained novice teachers achieve a 75% improvement in teaching ability, with these gains retained even after guidance is removed, and exhibit a 71% enhancement in applying skills beyond the training content.